Current evidence suggests that the pathogenesis of malignancy is a multistage phenomenon, that cancer results not from a single mutation but from a series of mutations. One of the best in vivo models for studying the progression of epithelial malignancies is the mouse two-state carcinogenesis system. In this system chemical carcinogen-initiated, TPA-promoted mouse skin tumors can be serially followed from benign premalignant papillomas to frank, invasive carcinomas. Recent evidence suggests that chemically-induced benign mouse skin papillomas harbor a point mutation in the c-Ha-ras oncogene suggesting that activation of this gene represents the initial step in the development of these skin tumors. However, subsequent genetic events in the development of these tumors, which might involve mutations of other (onco)genes, remain to be identified. We will approach this question by studying chemically-induced mouse skin carcinoma cell line that harbor double minute chromosome (DMs). These aberrant structures are frequently associated with gene amplification, and we plan to utilize fractionation techniques to molecularly clone amplified coding sequences from these cell lines. We will compare the structure of these cloned amplified sequences with other known genes by hybridization and DNA sequence analysis. Using Southern, Northern and slot-blot analysis, we will study the structure and expression of these sequences in normal epithelial cells as well as in the mouse skin tumors at different stages of malignant progression. If such sequences appear to be involved in the development of these skin tumors, then we will attempt to isolate homologous human sequences and determine if such human sequences might be involved in the development of human epithelial malignancies.